l e t t e r s

Mutations in SPRTN cause early onset hepatocellular carcinoma, genomic instability and progeroid features

Davor Lessel1,2,26, Bruno Vaz3,4,26, Swagata Halder3,4,26, Paul J Lockhart5,6,26, Ivana Marinovic-Terzic7,26, Jaime Lopez-Mosqueda8,9, Melanie Philipp10, Joe C H Sim5, Katherine R Smith11,12, Judith Oehler3,4, Elisa Cabrera13, Raimundo Freire13, Kate Pope5, Amsha Nahid11, Fiona Norris14, Richard J Leventer6,15,16, Martin B Delatycki5,6,17, Gotthold Barbi1, Simon von Ameln1, Josef Högel1, Marina Degoricija7, Regina Fertig4, Martin D Burkhalter18, Kay Hofmann19, Holger Thiele20, Janine Altmüller20, Gudrun Nürnberg20, Peter Nürnberg20–22, Melanie Bahlo10,23, George M Martin24, Cora M Aalfs25, Junko Oshima24, Janos Terzic7, Q1 5,6 8,9 3,4 1,2 David J Amor , Ivan Dikic , Kristijan Ramadan & Christian Kubisch 

Age-related degenerative and malignant diseases are major carcinogenesis and age-related diseases8,9. The latter are commonly challenges for health care systems. Elucidation of the molecular defined as segmental progeroid syndromes10 and can be caused mechanisms underlying carcinogenesis and age-associated by germline in encoding DNA repair pathologies is thus of growing biomedical relevance. We identified with concomitant cancer predisposition. Examples include WRN, biallelic germline mutations in SPRTN (also called C1orf124 or the Werner helicase , in Werner syndrome or BLM, the Bloom DVC1)1–7 in three patients from two unrelated families. All three helicase gene, in Bloom syndrome. In addition, mutations in nuclear patients are affected by a new segmental progeroid syndrome lamina–associated genes, for example, LMNA (encoding lamin A/C) characterized by genomic instability and susceptibility toward in Hutchinson-Gilford syndrome or BANF1 in Nestor-Guillermo early onset hepatocellular carcinoma. SPRTN was recently progeria11,12, can result in segmental progeria. Although LMNA proposed to have a function in translesional DNA synthesis mutations are also found in a few atypical cases of Werner syndrome13, and the prevention of mutagenesis1–7. Our in vivo and in vitro some patients with suspected Werner syndrome do not harbor characterization of identified mutations has uncovered an mutations in any known progeria gene14. essential role for SPRTN in the prevention of DNA replication Here we studied three patients from two unrelated families pre- stress during general DNA replication and in replication-related senting with early onset hepatocellular carcinoma (HCC), genomic G2/M-checkpoint regulation. In addition to demonstrating the instability and progeroid features. Consanguineous family A (Fig. 1a) pathogenicity of identified SPRTN mutations, our findings provide of Moroccan origin was referred to the International Registry of a molecular explanation of how SPRTN dysfunction causes Werner Syndrome, and the clinical characteristics of the affected boy accelerated aging and susceptibility toward carcinoma. in the family, A-IV:1, have been described previously15. The patient had short stature, bilateral cataracts, premature hair graying and Monogenic syndromes with highly penetrant tumor susceptibility died of HCC at the age of 17 years. Family B is a nonconsanguine- and/or signs of premature aging affecting more than one tissue have ous Australian family of European ancestry (Fig. 1b). Both affected been instrumental in identifying the genes and pathways involved in boys, B-II:1 and B-II:4, presented similar clinical features, including

1Institute of Human Genetics, University of Ulm, Ulm, Germany. 2Institute of Human Genetics, University Medical Center Hamburg-Eppendorf, Hamburg, Germany. 3Cancer Research UK and Medical Research Council Oxford Institute for Radiation Oncology, Department of Oncology, University of Oxford, Oxford, UK. 4Institute of Pharmacology and Toxicology, University of Zürich-Vetsuisse, Zürich, Switzerland. 5Bruce Lefroy Centre for Genetic Health Research, Murdoch Childrens Research Institute, Parkville, Victoria, Australia. 6Department of Paediatrics, The University of Melbourne, Parkville, Victoria, Australia. 7Department of Immunology and Medical Genetics, University of Split, School of Medicine, Split, Croatia. 8Buchmann Institute for Molecular Life Sciences, Goethe University, Frankfurt (Main), Germany. 9Institute of Biochemistry II, Goethe University School of Medicine, Frankfurt (Main), Germany. 10Department of Biochemistry and Molecular Biology, University of Ulm, Ulm, Germany. 11Bioinformatics Division, The Walter and Eliza Hall Institute, Parkville, Victoria, Australia. 12Department of Medical Biology, The University of Melbourne, Parkville, Victoria, Australia. 13Unidad de Investigación, Hospital Universitario de Canarias, Instituto de Tecnologías Biomédicas, La Laguna, Tenerife, Spain. 14Victorian Clinical Genetics Services, Murdoch Childrens Research Institute, Parkville, Victoria, Australia. 15Neuroscience Research, Murdoch Childrens Research Institute, Royal Children’s Hospital, Parkville, Victoria, Australia. 16Department of Neurology, Royal Children’s Hospital, Parkville, Victoria, Australia. 17Clinical Genetics, Austin Health, Heidelberg, Victoria, Australia. 18Leibniz Institute for Age Research, Fritz Lippmann Institute, Jena, Germany. 19Institute of Genetics, University of Cologne, Cologne, Germany. 20Cologne Center for Genomics, University of Cologne, Cologne, Germany. 21Center for Molecular Medicine Cologne, University of Cologne, Cologne, Germany. 22Cologne Excellence Cluster on Cellular Stress Responses in Aging-Associated Diseases, University of Cologne, Cologne, Germany. 23Department of Mathematics and Statistics, The University of Melbourne, Parkville, Victoria, Australia. 24Department of Pathology, University of Washington, Seattle, Washington, USA. 25Department of Clinical Genetics, Amsterdam Medical Centre, Amsterdam, the Netherlands. 26These authors contributed equally to this work. Correspondence should be addressed to C.K. ([email protected]), I.D. ([email protected]), D.J.A. ([email protected]), J.T. ([email protected]) or K.R. ([email protected]).

Received 24 April; accepted 4 September; published online XX XX 2014; doi:10.1038/ng.3103

Nature Genetics ADVANCE ONLINE PUBLICATION  l e t t e r s

Figure 1 Identification of causative SPRTN mutations. (a,b) The a Family A c pedigrees of families A and B. Filled and open symbols denote affected ? 1 2 and healthy individuals, respectively; an arrow indicates the index patient, I and diagonal lines indicate deceased status. The double line shows parental consanguinity, and the question mark denotes that the exact 1 2 degree of consanguinity is unknown. (c) Axial view of magnetic resonance II imaging of the liver of patient B-II:4. The green arrow indicates a 12 mm × 1 2 13 mm lesion mass with an absence of arterial phase enhancement within III segment VIII of the liver that was subsequently shown to be a HCC. (d) Analysis of total cell extracts of patients’ LCLs with SPRTN antibodies (Ab) raised against the N- or C-terminal part of the . (e) Genomic 1 2 IV localization and protein structure of SPRTN. The genomic structure is d based on the longest ORF containing five coding exons (black rectangles). kDa 75 CTRL B-II:1 B-II:4 The positions of the identified mutations are shown at both the gene (top) SPRTN b 50 and protein (bottom) levels. The protein diagram depicts the predicted 37 Family B functional domains of SPRTN. aa, amino acids. 1 25 ∆C-ter SPRTN

I N-terminal Ab 20 75 SPRTN low body weight, micrognathia, triangular face, muscular atrophy, 50 lipodystrophy, bilateral simian creases, delayed bone age and mild 1 2 3 4 37 II 25 joint restrictions in the fingers and elbows. Although hepatitis A, B 20 C-terminal Ab and C serologies and α-fetoprotein levels were normal in these two γ-tubulin boys, both developed early onset HCC at age 16 and 14, respectively e SPRTN (Fig. 1c). B-II:1 died at age 18 years from complications of acute fulminant hepatic failure. The clinical characteristics of all three Chr. 1 affected individuals are summarized and compared to those of known 1q42.2 segmental in Table 1. c.717-718+2delAGGT To identify the genetic cause of this putatively autosomal- c.350A>G c.721delA recessive segmental progeroid disorder, we performed genome-wide SPRTN linkage analysis (Supplementary Fig. 1) followed by exome sequenc- p.Tyr117Cys ing of unrelated individuals A-IV:1 and B-II:4. Bioinformatic filtering p.Lys239LysfsX7 identified SPRTN as the only gene with rare, biallelic mutations in p.Lys241AsnfsX8 the exomes of both individuals (Supplementary Tables 1 and 2). In 1 489 aa A-IV:1, a 1-bp deletion at cDNA position 721 bp (c.721delA) was the MIU SprT SHP PIP UBZ only nonannotated sequence change with a severe impact on protein structure within the homozygous regions and is predicted to intro- duce a premature stop codon at amino acid 249 (p.Lys241AsnfsX8). Werner syndrome but without mutations in WRN or LMNA14 revealed B-II:4 was compound heterozygous for a c.350A>G missense no other SPRTN , providing further evidence of extended alteration, resulting in the amino acid substitution p.Tyr117Cys, and locus heterogeneity for segmental progeroid syndromes. a 4-bp deletion at cDNA position 717 bp (c.717_718+2delAGGT). At We next performed morphological and immunohistochemical the cDNA level, this deletion predominantly caused intron inclusion, analyses of patients’ liver tumor biopsies. Staining with a C-terminal inducing a premature stop codon at amino acid 246 (p.Lys239LysfsX7). SPRTN antibody (Fig. 2a) showed the absence of the C-terminal part A very small fraction of cDNA demonstrated skipping of exon 4, of SPRTN in A-IV:1, thus confirming the truncation of the mutated resulting in a premature stop at position 161 bp (p.Val151IlefsX10) protein in A-IV:1 in vivo. We observed focal accumulations of anti- (Supplementary Fig. 2a–c). This finding was further supported by SPRTN immunoreactive material in B-II:1 and B-II:4, as well as in protein analysis, which identified a reduced amount of full-length idiopathic HCC. In vitro analysis of focal nuclear accumulation of protein and a new truncated protein (Fig. 1d). Sanger sequencing con- ectopically expressed wild-type (WT) SPRTN and mutant SPRTN firmed the mutations in all three patients (Supplementary Fig. 2d) from patients additionally supported the in vivo finding and dis- and cosegregation with disease state in their families (Supplementary closed that WT and p.Tyr117Cys SPRTN are able to form nuclear Table 3). None of these variants was present in dbSNP137 or the 1000 foci, but ∆C-ter SPRTN is not (Fig. 2b). Analyses of cancer biomark- Genomes data. The substitution p.Tyr117Cys is located in a puta- ers revealed strong focal accumulations of both γ-H2AX (H2AFX) tive zinc metalloprotease SprT domain five amino acids upstream of and 53BP1 (TP53BP1) (Fig. 2a). In addition, and opposite to what Glu112, which was recently shown to be necessary for the regulation we observed in patient primary cell lines (Fig. 3a–c), Ki-67 (MKI67) of error-prone translesional DNA synthesis (TLS)5. The identified staining in the biopsies of A-IV:1, B-II:1 and B-II:4 indicated a truncating mutations (∆C-ter SPRTN) lead to the loss of function- high proliferative index compared to healthy liver or idiopathic HCC ally important C terminal–located domains, including the ubiqui- (Fig. 2a). These data suggest a relatively aggressive neoplasm17,18. tin-segregase p97 (VCP)16-interacting motif (SHP), the proliferating We next tested whether the cellular phenotypes described previ- cell nuclear antigen interacting box (PIP) and the ubiquitin bind- ously in A-IV:1, namely chromosomal instability with concomitant ing domain (UBZ4; Fig. 1e). The C-terminal part of SPRTN has an sensitivity toward genotoxic agents and severe proliferation defects15, essential function at ultraviolet (UV)-induced stalled replication forks were also present in primary cell lines from B-II:1 and B-II:4. Indeed, by the removal of DNA polymerase η in a p97-dependent manner we found increased chromosomal instability in peripheral blood after the completion of TLS2,3. These genetic findings have already (Supplementary Fig. 3) and lymphoblastoid cell lines (LCLs) from provided strong evidence for the pathogenicity of the identified the patients, which was enhanced after treatment with mitomycin C mutations. The analysis of 48 additional patients with suspected (MMC) and 4-nitroquinoline 1-oxide (4-NQO) (Fig. 3d). In patient

 aDVANCE ONLINE PUBLICATION Nature Genetics l e t t e r s

Table 1 Clinical and cellular findings in Werner syndrome, atypical index revealed a severe growth defect in cultured B-II:1 fibroblasts Werner syndrome and patients (Fig. 3a–c), thus confirming prior findings in cells from A-IV:1 (ref. Atypical 15). Further, short interfering RNA (siRNA)-mediated depletion of Werner Werner SPRTN in HEK293T and U2OS human cell lines also caused chro- Clinical findings syndrome syndrome A-IV:1 B-II:1 B-II:4 mosomal instability and severe proliferation defects, respectively Short stature + + + − − (Supplementary Figs. 5 and 6). Low body weight + + + + + To assess the consequences of patient mutations in an in vivo Dermatological abnormalities + + − + − complementation assay, we performed morpholino-mediated down- Cataracts + − + − − regulation of the SPRTN ortholog in zebrafish (LOC101886162, Sparse hair + + − − − Premature hair graying + + + − − called here sprtn). sprtn silencing, verified by a GFP reporter assay γ Diabetes mellitus + + − − − (Supplementary Fig. 7a), led to a substantial increase of -H2AX Hypogonadism + − − − − foci (Supplementary Fig. 7b,c), indicating an evolutionarily con- Osteoporosis + − − − − served functional role of SPRTN in the DNA damage response. General skeletal abnormalities − − + − − When injecting higher doses of morpholino, embryos displayed Atherosclerosis + + − − − phenotypically normal development until the shield stage at 6 h Neoplasms + − + + + post fertilization (hpf), a stage when maternal gene products Hepatocellular carcinoma Rare − + + + are degraded20. At 10 hpf, however, embryos exhibited either early Micrognathia - + + + + mortality or were delayed in development up to 4 h. This growth Lipodystrophy + + ? + + retardation phenotype is compatible with the proliferation defects Muscular atrophy + + ? + + observed in patients’ fibroblasts and the relative growth deficits Attention deficit − − − + + observed in patients. This phenotype was partially rescued by Cellular findings co-injection of WT human SPRTN mRNA but not by equimolar Chromosomal instability + ? + + + amounts of mRNA from SPRTN harboring the identified mutations Sensitivity to genotoxic agents + ? + + + (Fig. 3e,f), thus further confirming their pathogenicity. Mutated gene WRN LMNA SPRTN SPRTN SPRTN The lack of sun sensitivity, severe chromosomal breakage, prolif- Atypical Werner syndrome includes cases with LMNA mutations. +, present in most eration defects and early onset HCC in patients indicated a more patients; −, not present in most patients; ?, not known. complex role of SPRTN in the maintenance of genome stability than solely TLS, as was proposed recently1–7. Defects in DNA replication fibroblasts, we observed multiple and variable aberrations that were have been proposed to be a major cause of the variegated translocation clonal in nature (Supplementary Fig. 4), which is compatible with mosaicism and genomic instability that consequently lead to aging variegated translocation mosaicism, a phenomenon previously and cancer21–23. To test for a role of SPRTN in DNA replication, we described in Werner syndrome cells19. Measurement of a proliferation analyzed the progression of replication forks directly by DNA fiber

a Ctrl HCC A-IV:1 B-II:1 B-II:4 b Flag-SPRTN Flag-SPRTN Flag-SPRTN (WT) (∆C-ter) (p.Tyr117Cys) Flag SPRTN (C-ter Ab)

γ-H2AX -H2AX γ M) + 1 h recovery µ

Merge 1 h CPT (1 1 53BP Ki-67

Figure 2 Severe DNA damage in hepatocellular carcinoma biopsies and focal nuclear accumulation of SPRTN. (a) Histological and immunohistochemical analyses of human liver biopsies from a healthy control (Ctrl), a patient with idiopathic, non–viral caused HCC and the HCC of patients with SPRTN mutations (A-IV:1, B-II:1 and B-II:4). The samples were stained with antibody raised against the C-terminal part of SPRTN (C-ter  Ab) or with antibodies against -H2AX, 53BP1 or Ki-67. The insets in the top three rows are at 3.5× magnification. (b) U2OS cells were transiently Q6 γ  transfected with Flag-tagged WT or mutant SPRTN and challenged with 1 µM of CPT to induce replication-related DSBs and thus mimic the DNA  damage observed in patients’ livers. Scale bars, 10 m (a,b). Q7 µ 

Nature Genetics ADVANCE ONLINE PUBLICATION  l e t t e r s

a b Ctrl B-II:1 c d Ctrl e ** Ctrl B-II:1 *** * B-II:1 Ki-67 100 B-II:4 *** *** *** 1.2 *** 100 20 80 *** *** Death 1.0 *** 80 Retarded

1,000) ** × 0.8 Unaffected 15 * 60 60 0.6 10 40 *** 40 0.4 *** DAPI *** Percentage

Growth rate 5 20 0.2 20 Ki-67–positive cells (%) 0 0 Mean aberrations/cell 0 0 (number of cells Untreated MMC 4-NQO – + – – – – –– Ctrl MO 2 4 6 8 Ctrl Time (d) B-II:1 – – + – – – – – Splice MO – – – + + + + + ATG MO – – – – + – – – WT SPRTN Figure 3 Genomic instability and cell proliferation defects. (a) A growth curve of control and – – – – – + – + ∆C-ter SPRTN – – – – – – + + p.Tyr117Cys patient primary skin fibroblasts. The data are from three independent experiments. (b) Primary skin SPRTN fibroblasts stained with the proliferation marker Ki-67. Scale bars, 10 µm. DAPI, 4′,6-diamidino-2- f + – – – –– Ctrl MO phenylindole. (c) Quantification of the data in b. The data are from three independent experiments – + + + + + ATG MO with greater than 100 cells scored per condition per experiment. (d) Chromosomal aberrations in – – – + – – WT SPRTN – – + – – + ∆C-ter SPRTN patients’ LCLs under untreated conditions or the following genotoxic conditions: 40 ng/ml of MMC p.Tyr117Cys kDa – – – – + + or 200 ng/ml of 4-NQO. The data are from three independent experiments with 33 analyzed SPRTN 100 Sprtn metaphase cells per condition per experiment. The data in a, c and d were analyzed with unpaired 70 t-test and are presented as the mean ± s.e.m. (e) Morpholino oligonucleotide (MO)-mediated gene SPRTN downregulation of Sprtn at 10 hpf and a rescue experiment with human WT SPRTN or SPRTN 55 harboring the two patient mutations. Ctrl MO denotes embryos injected with control MO, ATG MO 35 25 ∆C-ter SPRTN denotes embryos injected with SPRTN MO targeting the start codon, and splice MO denotes Actin embryos injected with a splice-site MO. The bar graph summarizes analyses in more than 100 embryos in three independent experiments showing the distribution of phenotypes observed after injections with 17.6 ng of MO and co-injections of 100 pg of SPRTN mRNA. P values in a and c–e were calculated using the χ2 test. *P < 0.01, **P < 0.001, ***P < 0.0001. (f) Western blot analysis of the experiment shown in e. Sprtn, endogenous zebrafish protein; SPRTN, ectopically expressed human protein. assay (Fig. 4a), the clearest method to unambiguously characterize after treatment with genotoxic agents that interfere with DNA the DNA replication machinery24. The speed of DNA replication replication, such as camptothecin (CPT) (Fig. 4h–j), a topoisomer- was significantly but only mildly affected in patients’ LCLs under ase I inhibitor that causes replication-related DSBs, or UV radia- unchallenged conditions (Fig. 4b), but increased levels of stalled forks tion (Supplementary Fig. 13). Notably, the G2/M checkpoint was (Fig. 4c) and newly fired origins (Fig. 4d) indicated DNA replica- completely functional when we created random and non–replication tion stress as the cause of DNA damage in these patients25. When we related DSBs using ionizing radiation (Supplementary Fig. 13). treated LCLs with a low dose of aphidicolin (APH), mimicking the The hypersensitivity of patient cells to replication-related genotoxic physiological barriers the DNA replication machinery approaches agents but not to ionizing radiation (Supplementary Fig. 14) during DNA synthesis26, patients’ cells showed a typical signature of correlates with G2/M leakage. DNA replication stress, namely shorter replication forks than in con- To further assess the role of patients’ mutations in DNA replication trol cells (Fig. 4e,f). This observation is comparable to cellular find- and G2/M-checkpoint regulation, we tested their function in U2OS ings in Werner and Bloom syndromes27,28. Moreover, B-II:1 fibroblasts cells that we depleted of endogenous SPRTN using siRNA (Fig. 5 showed markedly increased numbers of double-strand breaks (DSBs) and Supplementary Fig. 15). Ectopic expression of WT SPRTN in S-phase cells (Supplementary Fig. 8). To further confirm that restored DNA replication and G2/M-checkpoint defects. The expres- mutations in SPRTN are the cause of the DNA replication defect, sion of ∆C-ter SPRTN was also able to restore the progression of the we transfected patients’ cells with WT SPRTN, which almost com- DNA replication fork but to a much lesser extent then WT SPRTN, pletely corrected the replication defect (Fig. 4g and Supplementary suggesting that this mutant is defective in proper DNA synthesis Fig. 9a) and restored cellular proliferation (Supplementary Fig. 9b,c). (Fig. 5b). The difference in efficacy was even more pronounced when siRNA-mediated SPRTN depletion severely affected the progression we exposed cells to mild replication stress (APH treatment; Fig. 5c), of DNA replication and induced an increased number of stalled forks, further supporting the function of the C-terminal part of SPRTN in newly fired origins and formation of DSBs in the S phase of the cell TLS. Notably, the cells ectopically expressing p.Tyr117Cys SPRTN cycle (Supplementary Fig. 10), providing further evidence of the role were completely unable to restore DNA replication fork progression, of SPRTN in general DNA replication. Notably, depletion of DNA suggesting the essential role of the SprT domain in general DNA repli- polymerase η in patients’ cells or in SPRTN-depleted U2OS cells did cation. We obtained similar results in HEK293 cells (data not shown). not complement the DNA replication phenotype (Supplementary In contrast to DNA replication, cells expressing either of the patient Fig. 11), suggesting that DNA polymerase η is not the main substrate, mutations or coexpressing both mutations were equally defective in as reported previously2,3. activation of the G2/M checkpoint after exposure to UV radiation or To evaluate how replication-related DNA damage is transferred CPT treatment (Fig. 5d and data not shown), indicating the commu- to mitosis and may thus contribute to chromosomal instability, we nal function of the SprT domain and the C-terminal part of SPRTN measured the ability of patients’ cells to activate the G2/M check- in the regulation of the G2/M-checkpoint response. Taken together, point, which is the main guardian of genome stability after DNA rep- these data demonstrate that SPRTN dysfunction leads to sustained lication stress29. We exposed patients’ LCLs to different genotoxic DNA replication stress and consequent replication-related DNA dam- agents and measured the arrival of cells to mitosis by flow cytometry age, especially DSBs, which are transferred to the next cell generation (Fig. 4h–j and Supplementary Figs. 12 and 13). We observed a because of a leakage of the G2/M checkpoint and, consequently, lead severe G2/M-checkpoint defect in cells from B-II:1 and B-II:4 to cancer or aging.

 aDVANCE ONLINE PUBLICATION Nature Genetics l e t t e r s

IdU (+APH) a b CldU + IdU c d e f ** *** 15.0 *** ** 20 * 8 APH m)

m) 40 *** CldU (30 min) ldU (30 min) 10 CldU (30 min) ldU (30 min) µ 10.0 µ 10 Ongoing * ** 30 6 7.5 (1) replication fork Ctrl (2) Stalled forks 20 4 B-II:1 5.0 5 (3) 5 µm New origins 5 µm B-II:4 10 2 2.5 fork (tract length, fork (tract length,

Speed of DNA replication 0

Speed of DNA replication 0 0 0 Percentage of new origins Percentage of stalled forks Ctrl Ctrl Ctrl Ctrl B-II:1 B-II:4 B-II:1B-II:4 B-II:1B-II:4 B-II:1 B-II:4

g h i Ctrl B-II:1 B-II:4 j IdU (+APH) 104 Ctrl M: 49.7% M: 27.3% M: 26.2% B-II:1 3 m) *** *** 10 µ 10 B-II:4 1 µM CPT 102 0.6 8 16 h NOC

0 NOC 200 *** 6 400 101 *** 600 800 0.4 4 PI 1,000 G2: 18% G2: 18% G2: 16% h 100 104 2 M: 5.82% M: 11.4% M: 12.5%

fork (tract length, 3 0.2 0 10 Speed of DNA replication

eGFP-SPRTN (WT) – + – + – + 2 eGFP-empty: + – + – + – 10 M CPT + 16 0 and before DNA damage Ratio of mitotic cells after µ 0 101 Ctrl B-II:1 B-II:4 200 CPT + NOC recovery (NOC) 400 Ctrl 600 B-II:1 B-II:4

G2: 31% G2: 22% (leakage of G2/M-checkpoint arrest)

P-H3(S10) G2: 23% 1 h 800 0 PI 1,000 10 0 0 0 200 400 600 800 200 400 600 800 200 400 600 800 1,000 1,000 1,000 PI

Figure 4 DNA replication stress and leakage of the G2/M cell cycle checkpoint as the origin of genomic instability. (a) Schematic representation of a single DNA replication fork analysis by DNA fiber assay and its outcome ((1)–(3)). (b–d) DNA replication forks, as described in a, were analyzed in LCLs and quantified for velocity (b), percentage of stalled forks (c) and percentage of newly fired replication origins (d). n = 3; 100 CldU (5-chloro-2′- deoxyuridine) and 100 IdU (5-iodo-2′-deoxyuridine) symmetrical tracts were individually quantified, counted and are presented as total tract length (b), or 400 forks were randomly scored per condition per experiments (c,d). (e) Representative replication forks analyzed in LCLs when the incorporation of IdU was in the presence of a low dose of APH (0.1 µM). (f) The speed of replication forks under mild genotoxic conditions (IdU only). n = 3; 100 DNA fibers analyzed per experiment per condition. (g) LCLs were transfected with WT SPRTN, and the speed of replication fork was measured under mild genotoxic conditions, as in e. n = 2; 100 DNA fibers analyzed per experiment per condition. (h) Cell cycle profile of LCLs. NOC, nocodazole. PI, propidium iodide. (i) Mitotic (red triangle) LCLs without and with CPT treatment analyzed in the presence of NOC. (j) Quantification of i; n = 3. Data in b, f and g are shown as the median (bar) with the 25th–75th percentile range (box) and the 10th–90th percentile range (whiskers). Data in c, d and j are shown as the mean ± s.e.m. Data in b–d, f, g and j were analyzed with unpaired two-tailed t-test. *P < 0.0), **P < 0.001, ***P < 0.0001.

Studying the genetic and cellular basis of monogenic segmental diseases of the elderly8. HCC, although rarely occurring before the progeroid and tumor susceptibility syndromes has been a meaning- age of 40 years30, is the fifth most common malignancy and the third ful approach in unraveling the molecular mechanisms and pathways leading cause of cancer-related death worldwide31. Although major involved in the regulation of cancer development and common risk factors for HCC are well known, including hepatitis B and C

Figure 5 Characterization of patients’ mutations in a b CldU DNA replication and G2/M-checkpoint regulation. *** (a) Western blot analysis of U2OS cells depleted *** – + + + + + siSPRTN#1 NS of endogenous SPRTN by siRNA (siSPRTN#1) and 20 –– + –– – Flag-SPRTN (WT) *** m) simultaneously expressing siRNA-resistant WT A 15 –– – + – + Flag-SPTRN (p.Tyr117Cys) µ SPRTN, ∆C-ter SPRTN or p.Tyr117Cys SPRTN or kDa –– –– + + Flag-SPRTN (∆C-ter) 10 70 coexpressing of ∆C-ter SPRTN and p.Tyr117Cys SPRTN 55 5 SPRTN. (b,c) U2OS cells, as in a, labeled with 35 Speed of DN replication fork

CldU for 30 min (unchallenged conditions; ) or ∆C-ter SPRTN (tract length, b 25 0 – + + + + + siSPRTN#1 with IdU and treated with APH for 30 min (mild γ-tubulin –– + –– – Flag-SPRTN (WT) genotoxic conditions; ) were analyzed by DNA fiber c –– – – + + Flag-SPRTN (∆C-ter) assay. 1 µm of DNA tract length corresponds to 2.6 kb –– – + – + Flag-SPRTN (p.Tyr117Cys) of newly synthesized DNA32. n = 3; more than *** c IdU (+APH) d *** 100 DNA fibers analyzed per experiment and per *** condition. (d) U2OS cells, as in a, were analyzed for *** NS UV (10 J/m2) 15 s 0.6 the efficacy of the G2/M checkpoint after treatment m) A µ 10 ** ** with UV radiation, as described in Figure 4. The 10 ** 8 0.4 * graph summarizes three independent experiments. 6 NS The data in b and c are presented as the median 4 0.2 leakage) Speed of DN replication fork 2 (bar) with the 25th–75th percentile range (box) and UV treatment (tract length, after and before

0 (G2/M-checkpoint 0 the 10th–90th percentile range (whiskers). Data in – + + + + + siSPRTN#1 Ratio of mitotic cell – + + + + + siSPRTN#1 d are shown as a bar graph with the mean ± s.e.m. –– + –– – Flag-SPRTN (WT) –– + –– – Flag-SPRTN (WT) Data in b–d were analyzed with unpaired two-tailed –– – – + + Flag-SPRTN (∆C-ter) –– – – + + Flag-SPRTN (∆C-ter) –– – + – + Flag-SPRTN –– – + – + Flag-SPRTN t-test. *P < 0.05, **P < 0.001, ***P < 0.0001. (p.Tyr117Cys) (p.Tyr117Cys) NS, no significant difference between the groups. *** *** ***

Nature Genetics ADVANCE ONLINE PUBLICATION  l e t t e r s

infection and alcohol abuse, the molecular pathogenesis of HCC Reprints and permissions information is available online at http://www.nature.com/ remains largely elusive. All three patients presented here—in addition reprints/index.html. to showing signs of accelerated aging in selected tissues—developed early onset HCC, identifying SPRTN as a monogenic and apparently 1. Centore, R.C., Yazinski, S.A., Tse, A. & Zou, L. Spartan/C1orf124, a reader of PCNA ubiquitylation and a regulator of UV-induced DNA damage response. highly penetrant susceptibility gene for HCC. Consequently, our find- Mol. Cell 46, 625–635 (2012). ings suggest SPRTN as the subject of future studies of hepatocarcino- 2. Davis, E.J. et al. DVC1 (C1orf124) recruits the p97 protein segregase to sites of DNA damage. Nat. Struct. Mol. Biol. 19, 1093–1100 (2012). genesis and therapy. 3. Mosbech, A. et al. DVC1 (C1orf124) is a DNA damage–targeting p97 adaptor that promotes ubiquitin-dependent responses to replication blocks. Nat. Struct. Mol. URLs. The International Registry of Werner Syndrome, http://www. Biol. 19, 1084–1092 (2012). 4. Juhasz, S. et al. Characterization of human Spartan/C1orf124, an ubiquitin-PCNA wernersyndrome.org/registry/registry.html. interacting regulator of DNA damage tolerance. Nucleic Acids Res. 40, 10795–10808 (2012). Methods 5. Kim, M.S. et al. Regulation of error-prone translesion synthesis by Spartan/C1orf124. Nucleic Acids Res. 41, 1661–1668 (2013). Methods and any associated references are available in the online 6. Machida, Y., Kim, M.S. & Machida, Y.J. Spartan/C1orf124 is important to prevent version of the paper. UV-induced mutagenesis. Cell Cycle 11, 3395–3402 (2012). 7. Ghosal, G., Leung, J.W., Nair, B.C., Fong, K.W. & Chen, J. Proliferating cell nuclear antigen (PCNA)-binding protein C1orf124 is a regulator of translesion synthesis. Accession codes. NCBI reference sequences: UniGene: NM_032018, J. Biol. Chem. 287, 34225–34233 (2012). protein: NP_114407.3. Microarray probes: NM_032018.4. UniProt: 8. Burtner, C.R. & Kennedy, B.K. Progeria syndromes and ageing: what is the connection? Nat. Rev. Mol. Cell Biol. 11, 567–578 (2010). Q9H040. GenBank: AL512744. SPRTN mutations: NM_032018.5 9. Fletcher, O. & Houlston, R.S. Architecture of inherited susceptibility to common (c.721delA, ss1387933564, rs527236212; c.350A>G, ss1387933565, cancer. Nat. Rev. Cancer 10, 353–361 (2010).  Q2 rs527236213; c.717_718+2delAGGT, ss1387933572, rs587593493). 10. Martin, G.M. Genetic syndromes in man with potential relevance to the pathobiology  of aging. Birth Defects Orig. Artic. Ser. 14, 5–39 (1978). Q5  11. Navarro, C.L., Cau, P. & Levy, N. Molecular bases of progeroid syndromes. Note: Any Supplementary Information and Source Data files are available in the Hum. Mol. Genet. 15, R151–R161 (2006). online version of the paper. 12. Puente, X.S. et al. Exome sequencing and functional analysis identifies BANF1 mutation as the cause of a hereditary progeroid syndrome. Am. J. Hum. Genet. 88,  650–656 (2011). Q3 Ackno wledgments We are thankful to the family members for participation, G. Gillies for assistance 13. Chen, L. et al. LMNA mutations in atypical Werner’s syndrome. Lancet 362, 440–445 (2003). with patient samples, J. Schäfer for zebrafish care and Z. Garajova for technical 14. Oshima, J. & Hisama, F.M. Search and insights into novel genetic alterations leading assistance. We thank A. Raychaudhuri for initial help with the DNA fiber assay. to classical and atypical Werner syndrome. Gerontology 60, 239–246 (2014). We thank F. Böhm, Y. Böge and A. Weber from the University of Zurich and 15. Ruijs, M.W. et al. Atypical progeroid syndrome: an unknown helicase gene defect? L. Campo and K. Myers from the University of Oxford for providing healthy and Am. J. Med. Genet. A. 116A, 295–299 (2003). HCC human liver biopsies and performing histological and immunohistochemical 16. Vaz, B., Halder, S. & Ramadan, K. Role of p97/VCP (Cdc48) in genome stability. staining. The zebrafish γ-H2AX antibody was a kind gift of J. Amatruda (University Front. Genet. 4, 60 (2013). of Texas Southwestern). This work was supported by grants from Deutsche 17. King, K.L. et al. Ki-67 expression as a prognostic marker in patients with Forschungsgemeinschaft, the Cluster of Excellence ‘Macromolecular Complexes’ hepatocellular carcinoma. J. Gastroenterol. Hepatol. 13, 273–279 (1998). 18. Nowsheen, S., Aziz, K., Panayiotidis, M.I. & Georgakilas, A.G. Molecular of Goethe University Frankfurt (EXC115), the Landes-Offensive zur Entwicklung markers for cancer prognosis and treatment: have we struck gold? Cancer Lett. Wissenschaftlich-ökonomischer Exzellenz program Ubiquitin Networks of 327, 142–152 (2012). the State of Hesse, Germany and the European Research Council under the 19. Hoehn, H. et al. Variegated translocation mosaicism in human skin fibroblast European Union’s Seventh Framework Programme (FP7/2007-2013)/European cultures. Cytogenet. Cell Genet. 15, 282–298 (1975). Research Council grant agreement number 250241-LineUb to I.D., the European 20. Aanes, H. et al. Zebrafish mRNA sequencing deciphers novelties in transcriptome Commission (Marie Curie Reintegration Grant 268333 to M.P.), the Deutsche dynamics during maternal to zygotic transition. Genome Res. 21, 1328–1338 Stiftung für Herzforschung (M.P.), the Medical Research Council (MC_PC_ (2011). 12001/1) and the Swiss National Science Foundation (31003A_141197) to K.R., 21. Shen, J.C. & Loeb, L.A. The Werner syndrome gene: the molecular basis of RecQ helicase–deficiency diseases. Trends Genet. 16, 213–220 (2000). grants from the US National Institutes of Health (NIH) National Cancer Institute 22. Venkatesan, R.N. et al. Mutation at the polymerase active site of mouse (R24CA78088 and R24AG042328) to G.M.M., the NIH National Institute on Aging DNA polymerase δ increases genomic instability and accelerates tumorigenesis. Q4 (R21AG033313) to J. Oshima, the Ellison Medical Foundation to J. Oshima, the Mol. Cell. Biol. 27, 7669–7682 (2007). German Research Foundation (DFG) in the framework of the Cologne Excellence 23. Branzei, D. & Foiani, M. Maintaining genome stability at the replication fork. Cluster on Cellular Stress Responses in Aging-Associated Diseases to C.K., an Nat. Rev. Mol. Cell Biol. 11, 208–219 (2010). EMBO long-term fellowship to J.L.-M., a grant from the Croatian Ministry of 24. Zeman, M.K. & Cimprich, K.A. Causes and consequences of replication stress. Science, Education and Sport (216-0000000-3348) and a City of Split grant to J.T. Nat. Cell Biol. 16, 2–9 (2014). and I.D. K.R.S. is supported by a PhD scholarship funded by the Pratt Foundation. 25. Costantino, L. et al. Break-induced replication repair of damaged forks induces genomic duplications in human cells. Science 343, 88–91 (2014). M.B. is supported by an Australian Research Council Future Fellowship 26. Lukas, C. et al. 53BP1 nuclear bodies form around DNA lesions generated by (FT100100764). P.J.L. is supported by a National Health and Medical Research mitotic transmission of under replication stress. Nat. Cell Biol. 13, Council (NHMRC) Career Development Fellowship (APP1032364). This work was 243–253 (2011). made possible through Victorian State Government Operational Infrastructure 27. Sidorova, J.M., Li, N., Folch, A. & Monnat, R.J. Jr. The RecQ helicase WRN is Support and the Australian Government NHMRC Independent Research Institutes required for normal replication fork progression after DNA damage or replication Infrastructure Support Scheme. fork arrest. Cell Cycle 7, 796–807 (2008). 28. Davies, S.L., North, P.S. & Hickson, I.D. Role for BLM in replication-fork restart and suppression of origin firing after replicative stress. Nat. Struct. Mol. Biol. 14, AUTHOR CONTRIBUTIONS 677–679 (2007). D.L., B.V., S.H., P.J.L., I.M.-T., J.L.-M., M.P., J.C.H.S., K.R.S., J. Oehler, K.P., A.N., 29. Löbrich, M. & Jeggo, P.A. The impact of a negligent G2/M checkpoint on genomic F.N., R.J.L., M.B.D., G.B., S.v.A., J.H., M.D., R. Fertig, M.D.B., K.H., H.T., J.A., G.N., instability and cancer induction. Nat. Rev. Cancer 7, 861–869 (2007). P.N. and M.B. performed the experiments and did data analysis. E.C., R. Freire, 30. El-Serag, H.B. Hepatocellular carcinoma. N. Engl. J. Med. 365, 1118–1127 J. Oshima, G.M.M. and C.M.A. contributed materials and reagents used in the (2011). study. D.L., K.R. and C.K. wrote the manuscript. J.T., D.J.A., I.D., K.R. and 31. Caldwell, S. & Park, S.H. The epidemiology of hepatocellular cancer: from C.K. led and coordinated the entire project. the perspectives of public health problem to tumor biology. J. Gastroenterol. 44 (suppl. 19), 96–101 (2009). 32. Wilsker, D., Petermann, E., Helleday, T. & Bunz, F. Essential function of Chk1 can COMPETING FINANCIAL INTERESTS be uncoupled from DNA damage checkpoint and replication control. Proc. Natl. The authors declare no competing financial interests. Acad. Sci. USA 105, 20752–20757 (2008).

 aDVANCE ONLINE PUBLICATION Nature Genetics ONLINE METHODS with the Illumina TruSeq capture array. Novoalign (V2.08.03) alignment reads Ethical approval and study procedures. The International Registry of Werner mapped 57,905,140 reads uniquely to the hg19 reference genome. Presumed Syndrome has been recruiting patients suspected of having Werner syndrome PCR and optical duplicates were removed using Picard 1.65, and local rea- since 1988. Studies of family A and B were approved by the University of lignment was performed using RealignerTargetCreator and IndelRealigner Washington Institutional Review Board and the Human Research Ethics walkers from the Genome Analysis Toolkit (GATK)37,38. Greater than 80% Committee of the Royal Children’s Hospital, Melbourne, Australia, respec- of target sequences were covered at least tenfold, and the median coverage of tively. DNA samples from whole blood were isolated by standard procedures targeted bases was 40×. Single-nucleotide variants and small indel variants after written informed consent of participating individuals. were detected and genotyped using the UnifiedGenotyper walker from GATK version v2.3-3-g4706074. Variants were annotated using ANNOVAR39 against Linkage analysis in family A. Linkage analysis was performed using the RefSeq gene annotation; dbSNP build 137; 69 genomes from Complete Genome-Wide Human SNP Array 6.0. (Affymetrix, Inc., Santa Clara, CA, Genomics40; 1,092 genomes from the 1000 Genomes project, February 2012 USA). Data handling, evaluation and statistical analyses have been described release; 6,503 European and African American ancestry exomes from the in detail before33. National Heart, Lung and Blood Institute (NHLBI) Exome Sequencing Project (ESP6500, https://esp.gs.washington.edu/drupal/); and an in-house database of Linkage analysis in family B. All individuals in pedigree B were genotyped 131 exomes of various ethnicities. Read-backed phasing was performed using using Illumina Human610-Quad BeadChips (USA) at the Australian Genome HapCUT41. Variants located within the 33 linkage peaks were extracted and Research Facility, Melbourne. Genotypes were called using the GenCall algo- initially filtered for quality (phred scaled quality ≥13) and rarity (alternate rithm implemented in Illumina’s BeadStudio package. The LINKDATAGEN allele frequency of ≤1% in the 1000 Genomes and NHLBI ESP6500 data sets script34 was used to select a subset of 11,913 SNP markers for analysis. These and ≤4% in the Complete Genomics 69 and in-house data sets). We focused markers were chosen to be in approximate linkage equilibrium (spaced at least our analysis on rare homozygous or compound heterozygous variants pre- 0.15 cM apart) and to have high heterozygosity in the HapMap population dicted to affect protein sequence or splicing. Pairs of heterozygous variants of Utah residents with ancestry from northern and western Europe (CEU). located in the same gene that were inferred to be in cis phase by HapCUT41 Parametric linkage analysis was performed by MERLIN35 under a fully pen- were eliminated. etrant recessive inheritance model with a 0% phenocopy rate and a disease allele frequency of 0.00001. Allele frequencies from CEU were used. RT-PCR analysis. RT-PCR sequencing was done as described previously42,43. To distinguish between two SPRTN isoforms, we performed two RT-PCR reactions. Estimation of inbreeding coefficients in family B. FEstim36 was used to esti- Primer sequences used for reaction A to amplify the canonical 489-aa isoform mate genomic inbreeding coefficients for the genotyped members of family B. 1 of SPRTN spanning exons 2 through 5 were SPRTN-2F and SPRTN-5R, A subset of 11,374 high-heterozygosity markers in approximate linkage equi- with an expected WT size of 1,205 bp. Primer sequences used for reaction B librium were selected for analysis using LINKDATAGEN. FEstim was run in an to amplify the 250-aa isoform 2 of SPRTN spanning exons 2 through 4 were independent model with starting values F = 0.05 and A = 0.05. All inbreeding SPRTN-2F and intronic SPRTN-4r, with an expected WT size of 640 bp. coefficients were estimated to be 0.000, indicating no evidence of inbreeding. DNA fiber assay. The DNA fiber assay was performed as described previ- Candidate gene sequencing. For candidate gene analysis, we designed intronic ously44. Briefly, asynchronous LCL, U2OS or HEK293 cells were labeled with primers to PCR amplify coding exons and the respective exon-intron bounda- 30 µM of CldU for 30 min, washed three times with warm PBS and then ries by using genomic DNA of the affected individual. Primer pairs for the labeled with 250 µM of IdU for an additional 30 min. The reaction was termi- amplification of the five SPRTN coding exons and their approximately 50 bp nated by treating the cells with ice-cold PBS. Cells were lysed, and DNA fibers of flanking intronic sequences (RefSeq accession NM_032018.4) are available were spread onto glass slides, fixed with methanol and acetic acid, denatured on request. PCR products were sequenced on an ABI 3730 DNA Analyzer with with HCl, blocked with 2% BSA and stained with anti-rat and anti-mouse BigDye chemistry v3.1 (Applied Biosystems). Sequence traces were assem- 5-bromo-2′-deoxyuridine (BrdU) that specifically recognize either CldU bled, aligned and analyzed with SeqMan software (DNASTAR Lasergene). (Sigma, C6891) or IdU (Sigma, 17125). Anti-rat Cy3 (dilution 1:300, Jackson Cosegregation of the mutations in families A and B was tested by sequencing ImmunoResearch, 712-116-153) and anti-mouse Alexa-488 (dilution 1:300, the PCR product of exon 5, or of exons 3 and 4, respectively, amplified from Molecular Probes, A11001) were used as the respective secondary antibodies. genomic DNA of all participating family members. Microscopy was done using a Leica DMRB microscope with a DFC360FX camera. The lengths of the CldU- and IdU-labeled tracts were measured by Exome sequencing in A-IV:1. We sequenced the exome of the proband on two ImageJ software. Statistical analysis was done by GraphPad Prism software lanes of an Illumina GAIIx Sequencer using a single-read 150-bp protocol after using unpaired t-test. For the DNA fiber assay under genotoxic stress, the enrichment of exonic and splice-site sequences with the Agilent SureSelect second nucleotide (IdU) was incubated in the presence of 0.1µM APH. Human All Exon 50 Mb kit. We mapped >194 million reads to the hg19 human reference genome. Approximately 89% of target sequences were covered at Flow cytometry and G2/M-checkpoint assay. These analyses were performed as least 10-fold and 83% were covered at least 30-fold, with a mean coverage described previously45. In brief, cells were harvested, washed with PBS and subse- of about 112×. Data analysis of filter-passed reads was performed with the quently fixed in 3.6% formaldehyde at room temperature for 15 min. After wash- in-house pipeline V1.3 using BWA-short in combination with SAMTOOLS ing, the cells were permeabilized and blocked with 1% FBS and 0.1% saponin pileup 0.1.7 for the detection of SNPs and short insertions and deletions in PBS for 30 min. For 5-ethynyl-2′-deoxyuridine (EdU) analysis, cells were (indels). In-house–developed scripts were applied to detect protein changes, incubated with 10 µM EdU before harvesting. EdU was detected with a Click- affected splice sites and overlaps to known variations, with filtering against iT EdU Cell Proliferation Assay kit (Invitrogen, C10632). Alexa 647 ( 3458)– or dbSNP build 137, the 1000 Genomes Project data build February 2012 and Alexa 488 (9708)–conjugated anti–phosphorylated histone H3 (Ser10) (mitotic our in-house database of exome variants (with data from >200 exomes of marker) were used (dilution 1:100, Cell Signaling). Mitotic index was determined individuals affected by different disorders). We focused our analysis on rare as the ratio between the numbers of mitotic cells in the presence of nocodazole missense, nonsense, frameshift and splice-site mutations. The criteria for (400 nM for 16 h) after UV treatment compared to those in untreated cells. DNA a variation to be taken into account were as follows: >6 reads, phred scaled content was analyzed by DAPI. Cells were analyzed on a Beckman Coulter CyAn quality score >15, population allele frequency <1%, <10 times seen in our ADP Analyzer. A minimum of 10,000 events were counted. in-house database and >15% of the reads supporting the allele. Growth assay. 50,000 primary skin fibroblast or U2OS cells were seeded Exome sequencing in B-II:4. The exome of the surviving affected male B-II:4 at day 0. Every 24 h, cells were washed, trypsinized, resuspended in 1 ml was sequenced by Axeq Technologies (Korea) on three lanes of an Illumina medium (DMEM, Sigma, D6429) and counted (TC10 automated cell HiSeq sequencer using a paired-end 100-bp read protocol after exome capture counter, Bio-Rad).

doi:10.1038/ng.3103 Nature Genetics Histology. Human liver biopsy specimens were obtained from University otherwise stated. The numbers of breaks per cell were contrasted by two- Hospital Zürich, Academic Medical Centre Amsterdam and Royal Children’s sample Poisson tests, where P values were approximated using the χ2 Hospital Parkville. Biopsy specimens were registered in respective biobanks distribution with 1 degree of freedom. and kept anonymous. The study protocol was in accordance with the ethical guidelines of the Helsinki declaration. Liver samples were prepared from paraf- Cloning of LOC101886162, the SPRTN ortholog in zebrafish, called fin blocks according to standard histological protocols and hemalaun-eosin here sprtn. The zebrafish sprtn ortholog was cloned using a reciprocal, tblastn stained, or immunohistochemical staining was performed using Leica Bond protein query, BLAST (Basic Local Alignment Search Tool), which identified automated staining system. a partial sequence of a single zebrafish sprtn ortholog. Using 3′ RACE PCRs (First choice RLM RACE, Ambion, Life Technologies), the C-terminal part Immunofluorescence studies. Cells were grown on glass coverslips, fixed with and the 3′ UTR of SPRTN were obtained. The complete ORF of SPRTN was 4% formaldehyde, permeabilized with 0.2% Triton X-100, blocked with 5% BSA subsequently amplified from the cDNA of eight somite-stage zebrafish and in PBS and immunostained with the respective antibodies. Images of immunos- cloned into the pCS2+ vector. tained cells were taken with an epifluorescent microscope (Olympus BX51) and acquired with a charge-coupled device (CCD) camera (Orca AG), a Zeiss LSM Zebrafish maintenance and manipulation. Zebrafish were kept under con- 510 META laser scanning microscope or an SP2 Leica confocal microscope. trolled water and temperature conditions in a 14-h light and 10-h dark cycle. Fertilized eggs were allowed to develop at 28.5 °C up to the required stages and Antibodies. The following antibodies were used in this study: anti-SPRTN analyzed and processed as indicated. All husbandry procedures and experi- (rabbit, polyclonal) raised against the N-terminal part (1–240 aa) of SPRTN ments were approved by the ethics committee and research commission of the (dilution 1:1,000, home made); anti-SPRTN raised against the C-terminal part University of Ulm, Germany. For knockdown experiments, fertilized eggs were of SPRTN (dilution 1:1,000, Atlas, HPA 025073); anti–DNA polymerase η injected with RNA antisense MO or capped RNA transcribed with the mMes- (dilution 1:1,000, Abcam, Ab17725); anti–phosphorylated γ-H2AX (Ser139) sage mMachine Kit (Ambion) starting from linearized plasmids. Injections (dilution 1:300, Millipore, 05-636); anti-zebrafish γ-H2AX (dilution 1:1,000, were carried out at the one- to two-cell stage with an Eppendorf Femtojet gift of J. Amatruda); anti-rabbit 53BP1 (dilution 1:200, Santa Cruz, sc-22760); Microinjector (Germany). An antisense MO targeting the start codon of anti–Ki-67 (dilution 1:200, Millipore, MAB 4190); anti-rat BrdU (dilution SPRTN and a splice-site MO were used to generate loss-of-function zebrafish. 1:500, Abcam, 6326); anti-mouse BrdU (dilution 1:100, Becton Dickinson, Injections were controlled against those with a five-base mismatch control 347850); anti-mouse Alexa Flour 488 (dilution 1:300, Invitrogen, A21202), (Ctrl) MO. 1.5–17.6 ng of antisense MO against zebrafish SPRTN or Ctrl MO anti-mouse Alexa Flour 594 (dilution 1:300, Invitrogen, A11020); anti- were injected. The coding sequence of human SPRTN was amplified from mouse horseradish peroxidase (HRP) (dilution 1:10,000, Sigma, A2304); and pFlag–CMV-1–SPRTN and cloned into the pCRII-TOPO vector (Invitrogen, anti-rabbit HRP (1:10,000, Sigma, A0545). Darmstadt, Germany) and subsequently subcloned into vector pCS2+ by the use of BamHI and XhoI restriction sites followed by T4 ligation. The clini- Cell lines. Primary skin fibroblasts, Epstein-Barr virus (EBV)-transformed cal mutations were introduced using the QuikChange II XL Site-Directed LCLs, U2OS and HEK293T cells were used in this study. For stable transfected Mutagenesis Kit (Agilent, Waldbronn, Germany) and mutagenesis primers. GFP-SPRTN (WT) or GFP–empty vector control or patient LCLs, cells were Staining with antibody against phosphorylated zebrafish γ-H2AX (a gift from transfected by electroporation as indicated below and selected in a medium J. Amatruda) followed a standard protocol as described before46. containing G418/Geneticin (Gibco, 10131-027). GFP reporter assay. To verify the efficacy and specificity of MO-induced DNA primers and siRNA sequences. The DNA primers and siRNA sequences knockdown, capped mRNA of GFP fused to the whole ORF of SPRTN and used are listed in Supplementary Tables 4 and 5. parts of the 5′ UTR was injected alone or along with MO or CtrlL MO into zebrafish eggs at the one-cell stage. At 24 hpf, embryos were assayed for GFP Plasmids. The I.M.A.G.E. full-length SPRTN cDNA clone (IRATp970E1156D, fluorescence on a Keyence BZ8000K fluorescent microscope. ImaGenes) was cloned into pFlag–CMV-1 (Sigma), peGFP-C1 (Clontech) or pcDNA3.1 (Invitrogen). Mutants were cloned using PCR amplification and restriction enzyme digestion and recombination. Site-directed mutagenesis 33. Borck, G. et al. Loss-of-function mutations of ILDR1 cause autosomal-recessive was performed by PCR to introduce the desired mutations. The correctness hearing impairment DFNB42. Am. J. Hum. Genet. 88, 127–137 (2011). 34. Bahlo, M. & Bromhead, C.J. Generating linkage mapping files from Affymetrix SNP of the DNA sequence was verified by sequencing. chip data. Bioinformatics 25, 1961–1962 (2009). 35. Abecasis, G.R., Cherny, S.S., Cookson, W.O. & Cardon, L.R. Merlin—rapid analysis of Plasmid transfection. U2OS cells were transiently transfected with dense genetic maps using sparse gene flow trees. Nat. Genet. 30, 97–101 (2002). Lipofectamine 2000 (Life Technology), and primary skin fibroblasts and LCLs 36. Leutenegger, A.L. et al. Estimation of the inbreeding coefficient through use of genomic data. Am. J. Hum. Genet. 73, 516–523 (2003). were transfected by electroporation (Amaxa Nucleofactor Technology, Lonza) 37. DePristo, M.A. et al. A framework for variation discovery and genotyping using following the manufacturer’s instructions. next-generation DNA sequencing data. Nat. Genet. 43, 491–498 (2011). 38. McKenna, A. et al. The Genome Analysis Toolkit: a MapReduce framework for analyzing siRNA transfection. siRNA depletion experiments in mammalian cells were next-generation DNA sequencing data. Genome Res. 20, 1297–1303 (2010). 39. Wang, K., Li, M. & Hakonarson, H. ANNOVAR: functional annotation of genetic variants conducted using Lipofectamine RNAiMax (Invitrogen) according to the manu- from high-throughput sequencing data. Nucleic Acids Res. 38, e164 (2010). facturer’s instructions. A final concentration of 20 nM siRNA oligonucleotides 40. Drmanac, R. et al. sequencing using unchained base reads on was used. siRNA-transfected cells were analyzed 48 or 72 h after transfections. self-assembling DNA nanoarrays. Science 327, 78–81 (2010). 41. Bansal, V. & Bafna, V. HapCUT: an efficient and accurate algorithm for the haplotype assembly problem. Bioinformatics 24, i153–i159 (2008). analysis. For chromosome analysis, primary skin fibroblasts, 42. Hisama, F.M. et al. Coronary artery disease in a Werner syndrome–like form of LCLs and siRNA-transfected HEK293T cells were incubated with 40 ng/ml progeria characterized by low levels of progerin, a splice variant of lamin A. Am. MMC or 200 ng/ml 4-NQO or left untreated for an additional 24 h. Metaphase J. Med. Genet. A. 155A, 3002–3006 (2011). spreads and G banding were prepared using standard procedures, analyzed 43. Saha, B. et al. Ethnic-specific WRN mutations in South Asian Werner syndrome patients: potential founder effect in patients with Indian or Pakistani ancestry. using an Axio imaging 2 microscope (Zeiss, Jena, Germany) and captured Mol. Genet. Genomic Med. 1, 7–14 (2013). using Ikaros software (Metasystems, Altlussheim, Germany). 100 metaphase 44. Jackson, D.A. & Pombo, A. Replicon clusters are stable units of chromosome structure: spreads were scored for chromosomal aberrations in three independent experi- evidence that nuclear organization contributes to the efficient activation and ments. Based on 100 cells (untreated or treated with MMC or 4-NQO), the propagation of S phase in human cells. J. Cell Biol. 140, 1285–1295 (1998). 45. Cortez, D., Guntuku, S., Qin, J. & Elledge, S.J. ATR and ATRIP: partners in frequencies of aberrant cells were compared between LCLs from a healthy checkpoint signaling. Science 294, 1713–1716 (2001). individual (AG1010) and LCLs of both affected individuals or between SPRTN 46. Zhou, W. et al. FAN1 mutations cause karyomegalic interstitial nephritis, linking chronic siRNA treatment and a nonspecific siRNA using Fisher’s exact test, unless kidney failure to defective DNA damage repair. Nat. Genet. 44, 910–915 (2012).

Nature Genetics doi:10.1038/ng.3103 QUERY FORM

Nature Genetics

Manuscript ID [Art. Id: 3103] Author Editor

Publisher

AUTHOR: The following queries have arisen during the editing of your manuscript. Please answer queries by making the requisite cor- rections directly on the galley proof. It is also imperative that you include a typewritten list of all corrections and comments, as handwritten corrections sometimes cannot be read or are easily missed. Please verify receipt of proofs via e-mail

Query No. Nature of Query

Q1 Please carefully check the spelling and numbering of all author names and affiliations.

Q2 please ensure accession codes are schedule for release on or before this article’s scheduled publication date of 28 September 2014

Q3 Please check that all funders have been appropriately acknowledged and that all grant numbers are correct.

Q4 Please note, Nature Genetics style is to use the first initial and full spelled out last name of all authors with identical initials; we’ve edited accordingly.

Q5 Please note that we have edited and formatted the titles of Supplementary Figures to match the main text, as shown on the accompanying Integrated Supplementary Figures document. If you wish to make further changes to supplementary items, please notify us and send revised supplementary files (e.g., the Integrated Supplementary Figures document containing any revisions and/or a single PDF file containing any revised Supplementary Tables (unless these are to be posted as individual Excel files) and Notes). We will be unable to edit the PDF that you supply.

Q6 Correct magnification? If not, please edit to correct.

Q7 You indicated that there is no magnification in panel b, but it appears that the images at the very bottom of the figure (below the full-size micrographs) are magnified images of the boxed areas in the images directly above; if that is the case, please give the magnification of the bottom images.

Nature Publishing Group